Empirical models of plant demography are invaluable tools for predictive ecology and population management. Analyses of stage structured projection matrix models commonly study the long term growth rate of the population, predicted by the dominant eigenvalue of the projection matrix, and its sensitivity or elasticity to changes in demographic transition rates. Such analyses ignore the fact that natural populations are commonly disturbed away from stable stage structure. Disturbed stage structures suffer a period of transient amplification or attenuation before settling to stable growth or decline. Models of different populations or species of plants differ in predictions of “how big?” and “how soon?” these transient dynamics will be. Here we apply a range of established and new indices of transient population size (including reactivity, amplification envelopes and Kreiss bounds) to a database of projection matrix models of plant populations. In particular we ask whether predictions of transient amplification or attenuation are influenced by plant taxonomy, life history, or projection matrix size and structure.

Results/Conclusions

We find that predictions of transient amplification and attenuation correlate positively with long term population growth rate, and that transients differ between taxonomic subclasses. We explore links between life histories and transients. Finally, we reveal important correlations between predictions of transient amplification/attenuation and the size and structure of the projection matrices used to model plant populations. These findings emphasise the need to use the correct model structure for each plant’s life cycle if we wish to predict short-term as well as long-term population dynamics.